Modelagem do crescimento e produção de frutos da castanheira da Amazônia

• Autor(a) principal: Batista, Anderson Pedro Bernardina
• Data de Publicação: 2018
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: Repositório Institucional da UFLA
• Texto Completo: http://repositorio.ufla.br/jspui/handle/1/29037
• topic: Estatística espacial; Bertholletia excelsa; Castanha do Brasil; Produto florestal não madeireiro; Spatial statistics; Brazil nut; Non-timber product; Recursos Florestais e Engenharia Florestal

Knowledge of the natural regeneration and fruit yield of Bertholletia excelsa, as well as the factors and relationships that determine its variability, are fundamental for the management of non-timber resource production in natural forests. Thus, the aim of this thesis was to analyze aspects of autoecology and fruit yield of Bertholletia excelsa in different environments in the native forest in the Amazonia. The work was carried out at the Cajari River Extractive Reserve (Resex Cajari) located in the southern region of Amapá State, Brazil, with extension of 501,771 ha. The study of the growth projection of natural regeneration was carried out in different environments, non-flooded forest, cerrado-forest transition and secondary forest (capoeiras). All the diameters of the B. excelsa trees were measured to identify the rate of increase and then adjust the exponential model to estimate the growth trajectory. For the inventory of B. excelsa trees, four permanents plots of nine hectares (300 m x 300 m) in different environments, were distributed in native forest areas. For the study of tree neighborhood relationships with fruit yield of B. excelsa, all trees of other tree species with a diameter of 1.30 m (dbh) ≥ 3 cm were measured in the subplots of one permanent plot. The study of relationships with soil attributes, in which samples were collected one permanent plot. The relationships of the quantitative variables of the forest community in the surroundings of the trees of B. excelsa and attributes of the soil, was realized by geostatistical techniques. The variographic study was performed by experimental semivariogram and spatial theoretical models. The non-sampled points were estimated by ordinary kriging. The analysis of the spatial association of productive trees in different environments was performed by the bivariate K function. The explanatory modeling of fruit yield was performed using generalized linear models (GLM). It was observed that secondary forest (capoeira) presents favorable environment to the diametric growth of the natural regeneration of B. excelsa. The time to research the minimum diameter for fruit production was of 45 years, inferior than the other environments studied. It was observed that spatial association of species trees according to the yield classes has a different pattern depending on the different environments evaluated. The physicochemical attributes of the soil and tree vegetation were important to explain the variation of the fruit yield of the B. excelsa trees. The equation generated from GLM was important to understand the variables that influence the fruits yield of the B. excelsa trees in the native forest.